Blocking and triggering of plant immunity by Pseudomonas syringae effector AvrPto

Abstract

Bacterial effectors are double-edged swords that enhance bacterial virulence in susceptible plants while trigger resistance in plants carrying cognate resistance proteins. A well-known example of this is Pseudomonas syringae protein AvrPto that is delivered into plant cells through the type III secretion system. AvrPto inhibits immune responses in Arabidopsis plants but triggers resistance in some tomato plants carrying cognate resistance proteins Pto, a serine/threonine kinase, and Prf, a nucleotide-binding leucine-rich repeat protein. In a recent structural study we showed that AvrPto is an inhibitor of the Pto protein kinase. Because Pto closely resemble the kinase domain of receptor kinases, which include pattern recognition receptors (PRRs) crucial for plants to detect invading pathogens, we tested the possibility that PRRs such as FLS2 and EFR are targeted by AvrPto in susceptible plants. Indeed, AvrPto is capable of binding the FLS2 and EFR kinases to block plant immune responses when expressed in protoplasts. In Arabidopsis plants containing FLS2, the P. syringae strain lacking avrPto is compromised in its ability to multiply. However, the defect of the avrPto-deletion strain was alleviated in fls2 plants, indicating a role of AvrPto in overcoming FLS2-mediated resistance. Interestingly, the FLS2-AvrPto and Pto-AvrPto interactions share significant similarity, raising the tantalizing possibility that Pto has evolved as a molecular decoy of the intended targets of AvrPto.

Plants constantly face the threat of potential pathogens. The resistance or disease outcomes are, to large extent, determined by the interplay between the immune systems in the plant and virulence systems possessed by the pathogen. An important layer of plant immunity is PTI, which is activated when plants detect PAMPs originated from plant-associating microbes.1 This immunity is rendered ineffective when plants are challenged with virulent pathogens. For example, a number of P. syringae TTSS effectors are now known to inhibit plant PTI.2,3 Some plants contain NB-LRR proteins that directly or indirectly recognize some of the effectors to initiate ETI, a powerful defense system.4,5 Continued plant-pathogen co-evolution has enabled some pathogen to acquire effectors that specifically overcome ETI.5 Thus, resistance occurs when PTI or ETI is successfully activated, whereas disease ensues often when plant immunity is inhibited by pathogen effectors. Therefore, understanding how pathogen effectors function in inhibiting and triggering of plant defenses has been a major focus of the field.

PTI is initiated when PRRs recognize PAMPs at the plant cell surface.1 Plant PRRs known to detect pathogenic bacterial PAMPs include two receptor kinases, FLS2 and EFR. FLS2 recognizes a conserved N-terminal peptide of flagellin designated flg22, whereas EFR detects a conserved peptide from the elongation factor EF-Tu called elf18.6,7 The binding of ligands to FLS2 and EFR stimulates a cytoplasmic signaling pathway to trigger defenses. For example, the binding of flg22 to FLS2 induces its dimerization with another receptor-like kinase BAK1.8,9 Downstream, at least two MAP kinase cascades are activated to positively or negatively regulate PTI responses, which include transcriptional activation of gene expression, oxidative burst, and cell wall-based defenses.1

The Pseudomonas syringae effector AvrPto is directly recognized by the serine/threonine kinase Pto carried by some tomato plants to activate ETI specified by Prf, a NB-LRR protein.10,11 In Arabidopsis and susceptible tomato plants lacking Pto or Prf, AvrPto inhibits PTI and enhances bacterial virulence, but the virulence target(s) remained to be identified.12–14

In order to gain biochemical insights of the AvrPto protein, we solved the crystal structure of the Pto-AvrPto protein complex.15 Interestingly AvrPto binds Pto as a pseudosubstrate. In vitro analysis confirmed that AvrPto is an inhibitor of the Pto kinase. Sequence alignment and computer modeling suggested that Pto is structurally similar to the kinase domain of receptor kinases.16 We reasoned that AvrPto might have evolved as an inhibitor of receptor kinases. To test this possibility, we applied four different methods to determine if AvrPto is able to interact with FLS2 and EFR in vitro and in the protoplast. Protein pull-down and BIAcore surface plasmon resonance assays showed that AvrPto indeed interacts with the kinase domain of FLS2 and EFR in vitro. BiFC and co-immunoprecipitation (co-IP) assays showed that AvrPto is capable of interacting with the full-length FLS2 and EFR in the protoplast. Because FLS2 dimerizes with BAK1 when induced by flg22, the observed AvrPto-FLS2 interaction in the protoplast may be indirectly mediated by BAK1. To rule-out this possibility, we performed the co-IP assay in the absence of flg22 treatment and observed normal interaction. Furthermore, co-IP assays detected normal AvrPto-FLS2 interaction in bak1 null mutant protoplasts, indicating that the observed interaction is BAK1-independent. Together with the in vitro protein-protein interaction data, these results demonstrated that AvrPto can directly interact with FLS2 and EFR kinase domain in vivo.

If FLS2 is a major target of AvrPto, P. syringae bacteria lacking avrPto may display reduced virulence on Arabidopsis plants carrying FLS2. This deficiency should be alleviated in Arabidopsis plants lacking FLS2. Spray inoculation of wild-type (WT) and fls2 mutant Arabidopsis plants with the P. syringae strain DC3000, which carries avrPto, and the derivative strain lacking avrPto showed that this was indeed the case. The ΔavrPto strain grew to a lower level than did the DC3000 strain on WT Arabidopsis plants. On fls2 mutant plants, both strains grew to a similarly high level. AvrPto is similarly capable of interacting with the tomato flg22 receptor LeFLS2, and the ability to interact with LeFLS2 correlated with the virulence activity of AvrPto on susceptible tomato plants. These results demonstrated that the observed FLS2-AvrPto interaction is relevant to the immune suppression activity and virulence function of AvrPto.

We also tested the similarity of Pto-AvrPto and FLS2-AvrPto interactions. Both interactions required the AvrPto Y89 residue that makes direct contact with the kinase. The interaction also requires a functional ATP binding site, suggesting that autophosphorylation of the kinases are required for the recognition by AvrPto.15 The AvrPto I96 is required for interacting with Pto, and this residue is partially required for interaction with FLS2. Furthermore, Pto is capable of competing with FLS2 for binding to AvrPto. These results indicate that the Pto-AvrPto and FLS2-AvrPto interactions are at least partially similar and involve overlapping interaction surfaces, suggesting that Pto may be mimicking FLS2 for AvrPto recognition.

It has been proposed that Pto is a virulence target of AvrPto, and Prf acts by “guarding” the Pto protein (the guardee), allowing plant cells to sense the virulence activity of the bacterial effector and activate immune responses.17 This “guard model” has since been adopted to explain the recognition of a number of indirect recognitions between resistance proteins and pathogen effectors. However, the “guardees” identified to-date do not appear to fulfill a role of virulence targets. Pto is not required for the virulence function of AvrPto in tomato or Arabidopsis plants, indicating that other proteins are virulence targets.12 We propose an alternative model in which Pto has been adopted by plants to “deceive” the pathogen by mimicking the kinase domain of PRRs, the intended virulence targets for AvrPto (Fig. 1). The binding of AvrPto to Pto does not block PAMP-induced immune responses. Instead, it enables plant cells to sense the intruder and trigger ETI through Prf. Similarly, the tomato cysteine protease Rcr3, Arabidopsis RPM1-interacting protein RIN4, and Arabidopsis protein kinase PBS1 all mediate indirect recognitions between pathogen effectors and resistance proteins.18–22 None of these proteins is required for the virulence function of the corresponding pathogen effectors. In each case, there exist homologous host proteins that may be targeted by the effectors for virulence.18,23 There is a formal possibility that some of these NB-LRR proteins guard the decoy of virulence targets.

Abbreviations

PRRs	=	pattern recognition receptors
TTSS	=	type III secretion system
PAMPs	=	pathogen-associated molecular patterns
PTI	=	PAMP-triggered immunity
ETI	=	effector-triggered immunity
NB-LRR proteins	=	nucleotide-binding leucine-rich repeat proteins

Figures and Tables

Figure 1 Model for PTI-inhibition and ETI-induction by AvrPto. In susceptible plants, AvrPto binds PRRs including FLS2 and EFR. This blocks PTI activation and enhances host susceptibility to the bacterium. In resistant plants, Pto is evolved to mimic the kinase domain of PRRs and binds AvrPto to trigger ETI, resulting in a resistance outcome. The Pto-AvrPto structure is adopted from Xing et al.1 The blue rod connecting the LRR and kinase domains of FLS2 represents the transmembrane domain, whereas the rods on Pto and AvrPto denote myristic acid that targets these proteins to plasma membranes.

Addendum to: